Thursday, January 05, 2006

Human beings are Bayesian thinkers

This report does make sense to me. In particular, Steven Levitt, in his visit to Google pointed out that in the prehistoric world, the cost of making a false causual connection is low (such as carrying a rabbit's foot around), while the cost of not making an accurate prediction from a small sample size (i.e., you hear a roar of a tiger --- the last time you heard a roar of a tiger, your friend got eaten. Do you stop to think, "That's only a sample size of one?") could be very high, hence superstition prevails.

Indeed, some people suspect that the parsimony of Bayesian reasoning leads occasionally to it going spectacularly awry, with whatever process it is that forms the priors getting further and further off-track rather than converging on the correct distribution.

That might explain the emergence of superstitious behaviour, with an accidental correlation or two being misinterpreted by the brain as causal. A frequentist way of doing things would reduce the risk of that happening. But by the time the frequentist had enough data to draw a conclusion, he might already be dead.
Post a Comment